Minimization Of Expected Cost Research Articles

Reliability-based maintenance optimization of long-distance oil and gas transmission pipeline networks

This paper presents a maintenance model for optimal planning of pipeline network inspections subjected to corrosion and residual stress. The proposed model includes predictive degradation modeling, spatio-temporal reliability, and expected cost minimization. The pipeline networks are formed by linear segments and irregular zones such as elbows, weld joints, and flanges for putting pipes together and changing the direction of gas or liquid flow. The mean of Karhunen-Loève expansion is used to model space-variant corrosion, residual stress, and stochastic dependency in different exposure zones. Monte Carlo simulations are employed to compute the failure probability of individual components. Subsequently, a Bayesian network evaluates the failure probability of the complex pipeline network and identifies the most critical components. Depending on the objective or market demand uncertainties, a cost model is also proposed to optimize the inspection policy. The methodology is applied to a case study, and the results show that it can provide valuable insights to decision-makers to enhance complex systems' safety and reliability.

Bidding strategy design for electric vehicle aggregators in the day-ahead electricity market considering price volatility: A risk-averse approach

The uncertainties induced by electric vehicle (EV) demand and market operations pose huge challenges to the optimal bidding decision of the EV aggregator (EVA) in the day-ahead (DA) market. Note that a risk-neutral bidding solution with the expected cost minimization may make the EVA suffer a high financial loss in the market. As such, in this study, a risk-averse bidding strategy is developed to support the EVA to participate in the market via the conditional value-at-risk (CVaR) to handle market price volatility. Specifically, the strategy minimizes the CVaR metric over a collection of real-time (RT) clearing scenarios to reduce the energy transaction risk of the EVA in the market. Moreover, the model is reformulated as a linear programming (LP) problem that is mathematically tractable and can be efficiently solved. The proposed solution is extensively assessed through experiments based on the PJM market against the risk-neutral bidding strategy as a comparison benchmark. The numerical results reveal that the proposed risk-averse bidding strategy outperforms the risk-neutral one in terms of risk control, which enables the EVA to avoid suffering a huge financial loss incurred by RT clearing prices. In addition, the transformed LP model is superior to the original model in computational efficiency.

Active ordinal classification by querying relative information

Collecting and learning with auxiliary information is a way to further reduce the labeling cost of active learning. This paper studies the problem of active learning for ordinal classification by querying low-cost relative information (instance-pair relation information) through pairwise queries. Two challenges in this study that arise are how to train an ordinal classifier with absolute information (labeled data) and relative information simultaneously and how to select appropriate query pairs for querying. To solve the first problem, we convert the absolute and relative information into the class interval-labeled training instances form by introducing a class interval concept and two reasoning rules. Then, we design a new ordinal classification model for learning with the class interval-labeled training instances. For query pair selection, we specify that each query pair consists of an unlabeled instance and a labeled instance. The unlabeled instance is selected by a margin-based critical instance selection method, and the corresponding labeled instance is selected based on an expected cost minimization strategy. Extensive experiments on twelve public datasets validate that the proposed method is superior to the state-of-the-art methods.

Active Learning for Ordinal Classification Based on Adaptive Diversity-Based Uncertainty Sampling

Although a plethora of research has been conducted on active learning, little research attention has been focused on active learning for ordinal classification. Traditional multi-class active learning methods are typically designed for nominal multi-class classification. Therefore, they usually perform unsatisfactorily in ordinal classification settings. In ordinal classification, the cost of misclassifying an instance into an adjacent class is naturally lower than that of misclassifying it into a more disparate class. This principle is called ordering information. However, traditional active learning methods typically do not consider this ordering information during query selection. This paper proposes a novel adaptive hybrid active learning method for ordinal classification by considering the ordering information. In the proposed method, an uncertainty measure is introduced to select the hard-to-predict instances distributed between adjacent classes. In addition, a diversity measure is incorporated with the uncertainty measure to alleviate the potential sampling redundancy. Finally, an expected cost minimization measure with ordering information is designed. This measure balances the contributions of the uncertainty and diversity measures and prompts the algorithm to select the instances most likely to decrease the misclassification cost of the model. Extensive experiments on eleven public ordinal datasets demonstrate the superiority of the proposed method over several state-of-the-art methods.

Active learning for ordinal classification based on expected cost minimization

To date, a large number of active learning algorithms have been proposed, but active learning methods for ordinal classification are under-researched. For ordinal classification, there is a total ordering among the data classes, and it is natural that the cost of misclassifying an instance as an adjacent class should be lower than that of misclassifying it as a more disparate class. However, existing active learning algorithms typically do not consider the above ordering information in query selection. Thus, most of them do not perform satisfactorily in ordinal classification. This study proposes an active learning method for ordinal classification by considering the ordering information among classes. We design an expected cost minimization criterion that imbues the ordering information. Meanwhile, we incorporate it with an uncertainty sampling criterion to impose the query instance more informative. Furthermore, we introduce a candidate subset selection method based on the k-means algorithm to reduce the computational overhead led by the calculation of expected cost. Extensive experiments on nine public ordinal classification datasets demonstrate that the proposed method outperforms several baseline methods.

Pattern Selection Strategies for Pattern Databases in Probabilistic Planning

Recently, pattern databases have been extended to probabilistic planning, to derive heuristics for the objectives of goal probability maximization and expected cost minimization. While this approach yields both theoretical and practical advantages over techniques relying on determinization, the problem of selecting the patterns in the first place has only been scantily addressed as yet, through a method that systematically enumerates patterns up to a fixed size. Here we close this gap, extending pattern generation techniques known from classical planning to the probabilistic case. We consider hill-climbing as well as counter-example guided abstraction refinement (CEGAR) approaches, and show how they need to be adapted to obtain desired properties such as convergence to the perfect value function in the limit. Our experiments show substantial improvements over systematic pattern generation and the previous state of the art.

Optimal Replenishment Policies and Trade Credit for Integrated Inventory Problems in Fuzzy Environment

In this highly volatile, uncertain, complex, and ambiguous (VUCA) environment, a fuzzy inventory optimization problem is a hot topic that received the most attention from both academicians and practitioners. In this paper, the authors have studied the integrated inventory problems under trade credit in fuzzy environments, where the demand rate and deterioration rate are taken here as triangular fuzzy numbers. Accordingly, three types of fuzzy programming models as fuzzy expected cost minimization model, fuzzy α-cost minimization model, and credibility maximization model are built according to different management goals. Since the problem is shown to be NP-hard, this research focuses on proposing a hybrid intelligent algorithm by integrating fuzzy simulation, neural network, and genetic algorithm for solving these models. Then, some numerical examples are presented to illustrate the benefits of the models and evaluate the efficiency of the algorithms. The computational experiments show that the trade credit strategy has a positive impact on improving the supply chain’s performance and reducing its total cost. In addition, the maximum relative error of the objective values is less than 1%, which implies that the hybrid intelligent algorithm is robust to the parameter settings and effective in solving our models. Finally, the practical implications are discussed and useful managerial insights are given.

Matrix-geometric solution of multi-server queueing systems with Bernoulli scheduled modified vacation and retention of reneged customers: A meta-heuristic approach

ABSTRACT In the present article, we deal with the multi-server finite capacity queueing system with Bernoulli’s scheduled modified vacation policy and the realistic retaining policy of reneged customers. After completion of the service of any customer, the server decides whether to go for the vacation of random duration or to continue facilitating the service to the waiting customer, if any, present in the queue. All servers, homogeneous in nature, provide the state-dependent service following the threshold policy to reduce the overload of the system. The impatience behavior of customers like balking and reneging is also considered in the stochastic modeling of the studied problem. The matrix analytic approach is employed to obtain the steady-state probabilities with which various system performance measures are also developed with practical justification. Finally, the expected cost minimization problem is formulated and dealt with the meta-heuristic approach: particle swarm optimization (PSO). All results of numerical simulation and optimal analysis are summarized in tables and graphs to provide quick insight. The concluding remarks and future scopes have also been discussed.

Endogenous dynamic efficiency in the intertemporal optimization models of firm behavior

Existing methods for the measurement of technical efficiency in the dynamic production models obtain it from the implied distance functions without making use of the information about intertemporal economic behavior in the estimation beyond an indirect appeal to duality. The main limitation of such an estimation approach is that it does not allow for the dynamic evolution of efficiency that is explicitly optimized by the firm. This paper introduces a new conceptualization of efficiency that directly enters the firm’s intertemporal production decisions and is both explicitly costly and endogenously determined. We build a moment-based multiple-equation system estimation procedure that incorporates both the dynamic and static optimality conditions derived from the firm’s intertemporal expected cost minimization. We operationalize our methodology using a modified version of a Bayesian Exponentially Tilted Empirical Likelihood adjusted for the presence of dynamic latent variables in the model, which we showcase using the 1960–2004 U.S. agricultural farm production data. We find that allowing for potential endogenous adjustments in efficiency over time produces significantly higher estimates of technical efficiency, which is likely due to inherent inability of the more standard exogenous-efficiency model to properly credit firms for incurring efficiency-improvement adjustment costs. Our results also suggest material improvements in efficiency over time at an about 2.6% average annual rate, which contrasts with near-zero estimates of the exogenous efficiency change.

Endogenous Dynamic Efficiency in the Intertemporal Optimization Models of Firm Behavior

Existing methods for the measurement of technical efficiency in the dynamic production models obtain it from the implied distance functions without making use of the information about intertemporal economic behavior in the estimation beyond an indirect appeal to duality. The main limitation of such an estimation approach is that it does not allow for the dynamic evolution of efficiency that is explicitly optimized by the firm. This paper introduces a new conceptualization of efficiency that directly enters the firm's intertemporal production decisions and is both explicitly costly and endogenously determined. We build a moment-based multiple-equation system estimation procedure that incorporates both the dynamic and static optimality conditions derived from the firm's intertemporal expected cost minimization. We operationalize our methodology using a modified version of a Bayesian Exponentially Tilted Empirical Likelihood adjusted for the presence of dynamic latent variables in the model, which we showcase using the 1960-2004 U.S. agricultural farm production data. We find that allowing for potential endogenous adjustments in efficiency over time produces significantly higher estimates of technical efficiency, which is likely due to inherent inability of the more standard exogenous-efficiency model to properly credit firms for incurring efficiency-improvement adjustment costs. Our results also suggest material improvements in efficiency over time at an about 2.6% average annual rate, which contrasts with near-zero estimates of the exogenous efficiency change.

An IMPC Based Parking Assistance System

<div class="section abstract"><div class="htmlview paragraph">This paper summarizes progress and outcome from our research projects on adaptive model predictive control based parking management system, including parking motion planning and control strategy, as well as a localization strategy for parking spot. IMPC here refers to interactive model predictive control regime, which is characterized in that multiple agents implementing separate MPC strategy are incorporating information about their state, objective, and constraints. To predict future parking parameters, we proposed a practical framework which integrates anticipatory techniques with a model predictive approach that robustly models the uncertainty intrinsic of parking environment. The control regime puts comprehensive consideration into the interactions between vehicle subsystems, thus capable of optimizing trajectory under complex input disturbances under real-world driving scenarios. Moreover, Adaptive model predictive control is utilized to optimally minimize a cost function regarding time elapse, peripheral convenience and vehicle dynamics correspondent with surrounding vehicle states. The objective function is solved by dynamic programming under calibrated constraints in terms of velocity and steering accuracy, which yielded superior performance in comparison with convex programming. A dedicated guidance system was developed for navigating user to the parking spot for ease of locating parking location, which is characterized in that swift location and path are generated by BLE-based sensor fusion. Upon parking action finalization, the BLE sensor transmits the parking location to preferred communication equipment of user, which serves as goal of parking retrieve assignment. Simulation results show promising expected cost minimization in typical parking environments under consideration of numbers of trials, parking time and distance to destination. The achieved accuracy of MPC-led parking management strategy is able to improve the precision up to 0.3 meter, within the simulated highly disorganized parking environment. Meanwhile, the state of art park spot search module is able to shorten the time for drivers to locate their vehicle with positioning error of less than 1.2 meter.</div></div>

STOCHASTIC DISCRETE LOT-SIZING WITH LEAD TIMES FOR FUEL SUPPLY OPTIMIZATION

ABSTRACT We address the problem of expected cost minimization of meeting the uncertain fuel demand during a time planning horizon, where supply is provided by selecting discrete shipments with lead times. Due to uncertainty and the passage of time, corrective actions can be taken such as cancellation and postponement on supply of shipments with associated costs and delays. This problem is modeled as a stochastic multi-stage capacitated discrete lot-sizing problem with lead times. Computational experiments were performed on the resolution of various instances of the model for four information structures of uncertainty. The experimental optimal values and stochastic rating measures obtained show the validity and interest of the stochastic model, as well as the benefits that can be obtained with respect to a deterministic variant of the model that considers average demand.

New models of supply chain network design by different decision criteria under hybrid uncertainties

Supply chain network design is to determine factories and distribution centers to produce and distribute products, in order to satisfy the demand of customers to some extent. Supply chain network design plays an important strategic role in supply chain management and faces kinds of uncertainties, such as uncertain demand and cost. In order to deal with the impact of uncertain factors, this paper considers the supply chain network design under hybrid uncertainties, namely, the objective randomness of retailers’ demand and the subjective uncertainties of operating costs. We construct three models to handle managers’ different needs. They are the expected cost minimization model that minimizes the expected total cost, the $$\beta$$ -cost minimization model that minimizes the $$\beta$$ -cost which means that the chance measure of the actual cost not exceeding the cost is not less than the confidence level $$\beta$$ , and the chance measure maximization model that maximizes the chance measure of the actual cost not exceeding the given cost. This paper then transforms them into deterministic classes by uncertainty theory, and several numerical examples are presented to verify the validity of the models and algorithm.

Fixed Costs, Audit Production, and Audit Markets: Theory and Evidence

We analyze the role of discretionary fixed costs in audit production. Such costs are common (or joint) across a set of clients serviced by an accounting firm during a period. Our motivation is to better understand the implications of common discretionary fixed costs for the production and supply of audit services and audit market outcomes, such as the assurance level provided (audit quality) and audit costs and pricing, we conjecture that discretionary fixed costs are associated with the technology and other types of investments used by the auditor to transform audit effort (hours) into assurance for a portfolio of clients. We assume that such investments (and resulting costs) influence production primarily through the efficiency of audit effort (process improvements). We formally analyze the audit production problem in the context of auditor profit maximization and a more simple expected cost minimization formulation, and use examples, simulation methods, as well as explanatory figures to understand the implications of these models. Also, we use monotone comparative statics (Milgrom and Shannon [1994] to show the implications of client-specific losses (the auditor’s loss function) and the number of clients in an auditor’s portfolio on the level of fixed investments and audit assurance. The assumption that discretionary fixed costs are an important feature of audit service production yields a rich mix of interesting insights into the audit service market that are not found in the existing literature. Based on these analyses, we develop four testable hypotheses concerning the relations between audit quality and (1) the magnitude of client-specific possible losses from servicing a client; (2) the average client losses in an audit firm’s portfolio; (3) the number of clients in an audit firm’s portfolio; and (4) the variability of client losses in an audit firm’s portfolio. Using the absolute value of discretionary accruals as the proxy for audit quality and client size as the proxy for auditor losses, we find evidence consistent with the four hypotheses. In addition, using restatements as the measure of audit quality, we find some (weaker) evidence consistent with hypotheses 3 and 4.

Uncertain Solid Transportation Problem with Product Blending

In order to satisfy the required quality of demand points in a practical transportation system of chemical products such as petroleum and gasoline, blending of products should be considered. In this paper, we have considered a solid transportation problem with product blending in an uncertain environment, in which the fixed charges of vehicles and the unit transportation costs of products are assumed to be uncertain variables. Using different criteria to rank uncertain variables, we have constructed uncertain models for this problem, namely expected cost minimization model (ECMM) and chance-constrained uncertain model (CCUM). Within the framework of uncertainty theory, we prove that both of ECMM and CCUM can be transformed into the corresponding deterministic forms. Numerical examples are presented to illustrate the models, and the results are obtained by using the standard optimization solver Gurobi.

Freight transport platoon coordination and departure time scheduling under travel time uncertainty

The paper formulates and analyzes a freight transport platoon coordination and departure time scheduling problem under travel time uncertainty. The expected cost minimization framework accounts for travel time cost, schedule miss penalties and fuel cost. It is shown that platooning is beneficial only when scheduled arrival times differ less than a certain threshold. Travel time uncertainty typically reduces the threshold schedule difference for platooning to be beneficial. Platooning in networks is less beneficial on converging routes than diverging routes, due to delay at the merging point. The model provides valuable insights regarding platooning benefits for freight transport planning.

Fuzzy random programming models for location-allocation problem with applications

Three types of fuzzy random programming models based on the mean chance for the capacitated location-allocation problem with fuzzy random demands are proposed according to different criteria, including the expected cost minimization model, the α-cost minimization model, and the chance maximization model. In order to solve the proposed models, some hybrid intelligent algorithms are designed by integrating the network simplex algorithm, fuzzy random simulation, and genetic algorithm. Finally, some numerical examples about a container freight station problem are given to illustrate the effectiveness of the devised algorithms.

Risk-constrained framework for residential storage space heating load management

Residential demand response based on real time pricing provides a strong incentive for customers to reduce their energy payment. However, acting under an environment with time-varying prices will expose them to uncertain energy bills. This paper presents a risk-constrained framework for residential customers for scheduling the electric storage space heating load. The proposed decision framework attempts to accomplish desired settlement between expected cost minimization and cost deviation without altering the user's thermal comfort. The price and load uncertainty are captured by a scenario based stochastic programming approach. The optimization model is solved using Genetic Algorithm and implemented using a moving-window procedure. The simulation results demonstrate that the proposed framework for scheduling the storage space heating load provides a method to selectively hedge against the price and load uncertainty risk. The optimal framework will result in an improved interaction between the electrical aggregator and its customers under the smart grid paradigm.

A bandwidth sourcing and task allocation model in telecommunications under stochastic QoS guarantees

In this study, we model bandwidth provider selection and task allocation problem as an expected cost minimization problem with stochastic constraints. Two important parameters of quality of service (QoS), namely delay and jitter are considered as random variables to capture stochastic nature of telecom network environment. As solution methodology, stochastic model is converted into its deterministic equivalent and then a novel heuristic algorithm is proposed to solve resulting non-linear mixed integer programming model. We analyze how different prices, quality and task distribution affect the optimal behavior of the firm. Finally, performance of solution procedure is tested by several randomly generated scenarios and by a relaxation of this problem as tight lower bound.

Decision alternatives between expected cost minimization and worst case scenario in emergency supply – Second revision

Emergency supply chain management is often in the focal point of public scrutiny. Expected cost minimization may result in unacceptable high shortage for scenarios with small probabilities that is hardly tolerated by the public. To decrease the loss for the worst-case scenario, the min–max regret criterion could be applied for the pre-positioning decision. This alternative could get public support but results in high pre-positioning cost and large amount of unused resources since an unfavorable scenario with very low chance would dominate the solution. We suggest a compromising decision criterion and apply it for hurricane supply of commodities (like bottled water and canned food) in the Louisiana Gulf Coast. We coordinate the two stages, the preparedness and response decisions, in a large hierarchical emergency supply chain. The decision makers can utilize the major advantage of our parametric decision model providing a set of solution alternatives that dominate the decisions on different reliability levels.